Classifications

• • G—PHYSICS
  • G01—MEASURING; TESTING
  • G01T—MEASUREMENT OF NUCLEAR OR X-RADIATION
  • G01T1/00—Measuring X-radiation, gamma radiation, corpuscular radiation, or cosmic radiation
  • G01T1/16—Measuring radiation intensity
  • G01T1/20—Measuring radiation intensity with scintillation detectors
• • G—PHYSICS
  • G01—MEASURING; TESTING
  • G01T—MEASUREMENT OF NUCLEAR OR X-RADIATION
  • G01T1/00—Measuring X-radiation, gamma radiation, corpuscular radiation, or cosmic radiation
  • G01T1/16—Measuring radiation intensity
  • G01T1/20—Measuring radiation intensity with scintillation detectors
  • G01T1/201—Measuring radiation intensity with scintillation detectors using scintillating fibres
• • A—HUMAN NECESSITIES
  • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
  • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
  • A61B6/00—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment
  • A61B6/42—Arrangements for detecting radiation specially adapted for radiation diagnosis
  • A61B6/4208—Arrangements for detecting radiation specially adapted for radiation diagnosis characterised by using a particular type of detector
  • A61B6/4258—Arrangements for detecting radiation specially adapted for radiation diagnosis characterised by using a particular type of detector for detecting non x-ray radiation, e.g. gamma radiation
• • G—PHYSICS
  • G01—MEASURING; TESTING
  • G01T—MEASUREMENT OF NUCLEAR OR X-RADIATION
  • G01T1/00—Measuring X-radiation, gamma radiation, corpuscular radiation, or cosmic radiation
  • G01T1/16—Measuring radiation intensity
  • G01T1/20—Measuring radiation intensity with scintillation detectors
  • G01T1/208—Circuits specially adapted for scintillation detectors, e.g. for the photo-multiplier section
• • G—PHYSICS
  • G01—MEASURING; TESTING
  • G01T—MEASUREMENT OF NUCLEAR OR X-RADIATION
  • G01T1/00—Measuring X-radiation, gamma radiation, corpuscular radiation, or cosmic radiation
  • G01T1/16—Measuring radiation intensity
  • G01T1/24—Measuring radiation intensity with semiconductor detectors
  • G01T1/248—Silicon photomultipliers [SiPM], e.g. an avalanche photodiode [APD] array on a common Si substrate
• • G—PHYSICS
  • G01—MEASURING; TESTING
  • G01T—MEASUREMENT OF NUCLEAR OR X-RADIATION
  • G01T1/00—Measuring X-radiation, gamma radiation, corpuscular radiation, or cosmic radiation
  • G01T1/16—Measuring radiation intensity
  • G01T1/24—Measuring radiation intensity with semiconductor detectors
  • G01T1/249—Measuring radiation intensity with semiconductor detectors specially adapted for use in SPECT or PET
• • G—PHYSICS
  • G01—MEASURING; TESTING
  • G01T—MEASUREMENT OF NUCLEAR OR X-RADIATION
  • G01T1/00—Measuring X-radiation, gamma radiation, corpuscular radiation, or cosmic radiation
  • G01T1/29—Measurement performed on radiation beams, e.g. position or section of the beam; Measurement of spatial distribution of radiation
  • G01T1/2914—Measurement of spatial distribution of radiation
  • G01T1/2985—In depth localisation, e.g. using positron emitters; Tomographic imaging (longitudinal and transverse section imaging; apparatus for radiation diagnosis sequentially in different planes, steroscopic radiation diagnosis)

Definitions

• the present invention relates to three-dimensional computing.
• the present invention relates to proton-computed tomography.
• Generating an internal 3D image of an object with proton computed tomography starts by firing many protons through the object.
• Each proton's path i.e., trace history
• the collection of all the trace histories are used as input to a computer program that generates the image.
• Detectors for pCT in the prior art employ larger bulky plastic cube shaped tubes (such as in Pemler, et al.) with large and bulky photon sensors that require much larger volumes that present problems when mounted on a proton gantry with limited space around the patient.
• Pemler, et al. describes the possibility to perform radiography with protons and used a single X-Y plane in front of the object being analyzed and a single X-Y plane after.
• Pemler, et al. also uses vacuum photomultipliers, square fibers, and combinatorial readout. Amaldi, et al.
• the current pCT detector at Loma Linda uses silicon wafers that are limited in available area (9 cm x 9 cm), thus requiring a mosaic of overlapping tiles to achieve large area (27 x 36 cm).
• the maximum size now produced of silicon wafers is 10 cm x 10 cm.
• the tiles are shingle overlapped or placed with edges butting, which create dead space in the detector or double layers that require "mathematical removal" during image calculations.
• the present invention provides for a proton computed tomography (pCT) detector system, including two tracking (2D) detectors in sequence on a first side of an object to be imaged, two tracking (2D) detectors in sequence on an opposite side of the object to be imaged, and a calorimeter, wherein the tracking detectors include plastic scintillation fibers.
• pCT proton computed tomography
• the present invention also provides for a method of imaging an object by emitting protons from a source through two tracking detectors, through and around the object, and through two opposite tracking detectors, detecting energy of the protons with a calorimeter, and imaging the object.
• FIGURE 1 is a layout of the pCT detector system of the present invention
• FIGURE 2 is a three-dimensional representation of the pCT detector system core idea of the present invention
• FIGURES 3A-3C are three-dimensional representations of one plane of the tracking detector of the present invention.
• FIGURE 4 is a three-dimensional representation of the system of the present invention.
• FIGURE 5 is flow-chart showing the process of generating images with the scanner system
• FIGURE 6A is a front view of a silicon strip detector, and FIGURE 6B is a graph of a proton beam detected;
• FIGURE 7 is a perspective view of fibers of the tracking detector.
• FIGURE 8 is perspective view of a scintillating plate as part of the calorimeter.
• the present invention provides for a pCT detector system 10, shown generally in FIGURE 1 , for generating 3D images from proton trace histories.
• the pCT detector system 10 preferably includes four tracking detectors 12, a calorimeter 14, and a computer cluster (not shown).
• the tracking detectors 12 are used to detect the coordinates in the X and Y plane of the protons.
• Each detector 12 has two planes, an X plane and a Y plane, with the fibers (strips) oriented in the appropriate direction for X and Y measurements of the proton tracks.
• Each detector is fabricated by placing one X and one Y plane in close proximity with a thin foam-like material in between.
• the tracking detectors 12 include 1 mm diameter thin circular plastic scintillation fibers 22, closely packed in a row, and covering the full area of the imaging field, preferably 27 cm by 36 cm, or 18 cm x 36 cm.
• the tracking detectors 12 can be any other suitable size.
• the core of the fibers is preferably polystyrene with cladding of PMMA (Poly(methyl methacrylate)).
• the tracking detectors 12 also include appropriate electronics in order to detect the position of the protons and relay this data to a computer cluster.
• the tracking detectors 12 are arranged so that two are placed in sequence on a first side of an object 16 to be imaged and two are placed in sequence on an opposite side of the object 16, as shown in FIGURES 1 and 2.
• Scanning magnets (or a source) 18 that emit the proton beams of the system 10 are located at a position opposite to the calorimeter 14 on either side of the outer tracking detectors 2.
• the scanning magnets 18 can be located 240 cm from the center of the object 16, and the calorimeter 14 can be 65 cm from the center of the object 16.
• the inner tracking detectors 12 can be 15 cm from the center of the object 16 and the outer tracking detectors 12 can be 30 cm from the center of the object 16. Other distances can be used as appropriate.
• the tracking detectors 12 include the use of low cost Silicon Photomultipliers (SiPM) 20 attached to the plastic scintillating fibers 22 for signal amplification in lieu of phototubes as used in the prior art, as shown in FIGURES 3A-3C.
• the SiPMs 20 provide continuous coverage and remove the tile construction effect. SiPMs are unaffected by magnetics contrary to phototubes. SiPMs have been historically used in particle detectors. SiPMs are intrinsically fast, of order 100 nanosecond resolving time between events. Therefore, this design allows much higher data acquisition rates than previous detector systems.
• the plastic scintillating fibers 22 can have a cross-sectional shape of a square (shown in FIGURE 7), circle, or hexagon.
• the tracking detectors 12 with SiPMs 20 can include two scintillating fibers 22 simultaneously, alternatively, one fiber 22 can be used for one SiPM 20.
• the fiber 22 diameter is in the range of 0.75 mm - 1 mm.
• the fibers 22 can be polystyrene scintillating fibers that produce light when protons cross their thickness. Two layers of fibers 22 increase detection efficiency in contrast to one layer. Another orthogonal double layer of fibers 22 can be included in order to have a 2D coordinate of the protons in space.
• the tracking detectors can include a mechanical support 24 for the SiPM, as well as a Rohacell support 26 that serves as a support for scintillating fiber 22 placing. Any other supports can be used as known in the art in order to create the tracking detectors 12.
• the area of 27x36 cm 2 (tracking part) is covered. If the picked diameter of the scintillating fiber 22 is 1 mm then 270 fibers, 270 SiPMs, and 270 channels of electronics are needed. This is true for a one projection only (X).
• FIGURE 6A shows a side view of the tracking detector 12 of the present invention.
• this detector 12 there are two layers of silicon strips.
• the silicon sensors which connect to the SiPms 20 are 89.5 mm x 89.5 mm with a strip pitch of 238 ⁇ and 400 ⁇ thickness.
• the strips of each layer are individually connected to six ASICs * 64 strips.
• FIGURE 6B shows a view of a proton beam detected by the tracking detector 12 with X and Y coordinates.
• the calorimeter 14 is used to detect proton energy that is emitted from the source 18.
• SiPM 20 with a diameter of 1.2 mm can provide a readout (digital or analog) through 1.2 mm wavelength shifter (WLS) fibers 36.
• the SiPM 20 used in the calorimeter 14 can be the same type as used in the tracking detectors 12.
• the calorimeter 14 can be arranged in a stack (at least two) of 3 mm scintillator plates 34, as shown in FIGURE 8. In this case, the total number of channels are about 120. This is arrived at by the following.
• One scintillating plate 34 includes 1 WLS fiber and 1 SiPM.
• the total number of plates is 120. So, there are 120 plates, 120 WLS fibers, 120 SIPM, and 120 channels of electronics (120 channels).
• the calorimeter 14 can further include appropriate electronics to relay data to the computer cluster.
• FIGURE 4 shows the system 10 with a rotational stage 28 that holds and rotates an object 16 during proton exposition.
• the present invention provides for a method of imaging an object by emitting protons from a source through two tracking detectors, through and around the object, and though two opposite tracking detectors, detecting the protons with a calorimeter, and imaging the object.
• protons are emitted from a source 18 (scanning magnets) through two tracking detectors 12, through and around the object 16 to be imaged, through additional two tracking detectors 12, and finally pass through the energy detector (i.e., calorimeter 14) (FIGURE 1 ).
• Protons of sufficient energy can penetrate the human body and can be tracked upon entry and exit of the tracking detectors 12.
• the X and Y position of the protons are measured and detected on the X and Y planes of each of the tracking detectors 12.
• the system 10 includes 4 detectors or eight planes of these fiber trackers to record the position of each proton track as it passes through each plane.
• the object 16 being imaged is located such that two detectors 12 are on each side of the image.
• the detectors 12 rotate about the object 360 degrees during operation.
• Generating 3D proton computed tomography images requires data from a large number of proton histories to be stored in memory.
• Previous computer programs executed on stand-alone general purpose graphical processing unit (GPGPU) workstations implemented in prior art were constrained by the demand on computer memory.
• the approach described here is not limited to demands on computer memory in the same manner as the reconstruction is executed on multiple workstations simultaneously.
• the resulting data recorded from the detector and the calorimeter can be fed into an image reconstruction software program on a computer cluster to allow high-resolution images to be formed in under 10 minutes after irradiation.
• the computer program is written in terms of the Message Passing Interface (MP! standard.
• MPI Message Passing Interface
• Writing the program in terms of the MPI standard enables one to run the program on a cluster of many computers as opposed to a single computer.
• the program can be run on approximately 768 CPUs (computers) plus 96 GPUs (graphic processing units). In this way, it is possible to bring to bear the combined memory and computational power of many computers at one time, thus substantially reducing the time it takes to run the program (by orders of magnitude) as well as increasing the problem size (as measured by size of image space and the sharpness of the image) the program can image. Therefore, data acquired from the calorimeter 14 can be analyzed on a cluster of multiple computers with this standard.
• the compact memory representation enables one to deploy the computer program to solve the entire 3D image space at one time. This results in images of higher quality as it allows one to use the information from the trace histories of more protons as compared to the "slice and assemble" technique used in the prior art.
• the new computer program of the present invention is capable of producing high-quality 3D internal images for cancer patients being treated with proton therapy.
• the new computer program and specifically the two novel techniques of adding MPI and the compact memory representation, is able to produce quality images faster than any other known methods and, in fact, so fast (i.e., less than 10 minutes) that the program can not only be used to augment or replace X-ray generated images used in proton treatment of cancer today, but medical care providers can find new uses for such images to improve the quality of care delivered to the patient (e.g., just before treating the patient with proton therapy generate one more set of images to refine the treatment plan).
• pCT imaging and this new computer program can become the state of the art, and thus become the de facto standard, for generating images for every cancer patient globally treated with proton therapy.
• Example 1 The computer program is further described in Example 1 below, and an example of code is shown in EXAMPLE 2.
• a foreman computer distributes equal amounts of proton histories to multiple worker computers and the foreman computer itself.
• An Integrated Electron Density (I ED) and Most Likely Path (MLP) are computed by each of the computers.
• a solution vector is solved for and stored on computer readable memory in each of the computers. Copies of the solution vector from the worker computers are sent to the foreman computer and combined and stored on computer readable media.
• the combined solution vector is tested by the foreman computer, and if the combined solution vector is determined to be done, an image is then produced of the object. If it is determined that the combined solution vector is not done, the combined solution vector is transmitted to the worker computers, and each of the above steps are repeated until the combined solution vector is determined to be done by the foreman computer, and an image is produced of the object.
• FIGURE 5 shows a flow chart of the general process of the present invention.
• the proton detection process is shown in the bottom row, and data from the tracking detectors 12 and calorimeter 14 is analyzed from the top using parallel processors and the algorithms described above. Finally, an image is generated, which looks generally like a CT image.
• Our proposed solution to implement string averaging methods is to use MPI with set of worker processes where one worker is designated as the foreman.
• the foreman reads the proton histories and distributes them evenly across all the workers, saving an equal share for himself. From their portion of the histories each worker computes (one time only) IED and MLP and initializes their solution vector which represents the entire voxel space (i.e., each worker has its own copy of the entire solution).
• the program then enters an iteration loop in which:
• the foreman collects and combines with his own the solution vectors from all the other workers and tests combined solution vector. If combined solution is "done” , then foreman tells all the other workers to end. If "not done' ! the foreman broadcasts the combined solution vector to all the other workers who, in turn and along with the foreman, use that as their starting point for their next iteration.
• V T desired resolution of image e.g., 1 mm
• V s MLP image resolution oversample rate (e.g., 2)
• V p protons per voxel number of imaging protons travelling through each voxel
• Each proton history is characterized by an 11-tuple comprised of four ⁇ x, y > tuples (one for each detector plane), input and output energies (_3 ⁇ 4 ra and i5 ou t) , and a projection angle. These values require p h + b pw bytes of storage for each of the four ⁇ x, y > tuples plus three floats for the last three values.
• Each history is processed to determine the 3-D surface of the phantom (i.e., space carving).
• the voxel space is represented by an array, one element for each voxel, and the elements of the array are set or cleared if the phantom does nor does not occupy the voxel, respectively.
• the memory (I m ) required to store the processed input projection histories for each worker is
• IED Integrated Electron Density
• IED m H w f (2) MLP.
• MLP Most Likely Path
• chord length a floating point number
• V3 ⁇ 4, ⁇ , d the voxel cube dimensions
• d the firing angle
• a simple memory layout that "connects" a chord length and the angle that produced it to the voxel indices of the protons fired from that angle is to maintain a vector of chord lengths and two additional vectors in tandem as follows;
• MLP m I a f + H w ⁇ b Ia + p+ ⁇ l ⁇ + l
• solution vector is an array of floats whose length is determined by the number of voxels V .
• Vm the number of voxels
• Reccurrence forumla that keeps a running mean (M_k) and running SSE (S_k) for some stream of values X i
• M_i M_(i-1) + [(x_i - M_(i-l))/i]
• N N_l + N_2
• op_res [ i] .mean + (delta * op[i].n / combined_n);
• op_res[i].n + op[i].n;
• MPI_Comm_size MPI_COMM_WORLD, Snumprocs
• MPI_Comm_rank ( MPI_COMM_WORLD , &my_id ) ;
• src_data[i] (random) ) % RandomizedRange ) - (RandomizedRange/2 ) ; printf ( "my_id %d: sample_data %f ⁇ n" , my_id, src_data[i] ) ;
• N sizeof (SampleData) /sizeof (float) ;
• MPI_Op_create (MPI_JJser_function* ) combine_stats , // my user-defined func
• chord lengths */ float get_chord_length(int h_idx);
• TRACK_DX ( RANGEJTRACK_X/MAX_TRACK_X_ID ) ;
• TRACK_DY (RANGE_TRACK_Y/MAX_TRACK_Y_ID ) ;
• the data reconstruction space has two coordinate systems. */
• the first is measured in centimeters and has the same XYZ orientation */
• the second coordinate system is based on a plane-major, row major ordering */ /* scheme with planes parallel to the beam rotation plane.
• /* is parallel to X. Row axis is parallel to Z. Plane axis is parallel to Y. */
• the reconstruction volume is an array a[MAXJPLANE ] [ AX_ROW] [MAX_COL] */
• MAX_DAT [ 3 ] ⁇ MAX_COLS , MAX_ROWS, MAX_PLANES ⁇ ;
• VOLUME_ _MAX_ z VOLUME_MIN_Z + MAX_ _ROWS * D_ _ROW;
• NICK VOXEL_CUBE_DIAG_VOXEL_LEN and NVOXELS (later) need to be corrected // NICK: when MAX_ ⁇ ROWS , COLS,PLANES ⁇ are convereted from nvoxels (which is // NICK: what they are now) to cm ... VOXEL_CUBE_DIAG_VOXEL_LEN and
• const float VOXEL_CUBE_DIAG_VOXEL__LEN sqrt ( (MAX_ROWS*MAX ROWS)
• NVOXELS MAX_COLS*MAX_ROWS*MAX_PLANES ;
• dst[ 0 ] (dst[0] - VOLUME_MIN_ ) / D_COL ;
• dst[ 1 ] (dst[ 1] - VOLUME_MIN_Z ) / D_RO ;
• dst[2] (dst[2] - VOLUME_MIN_Y ) / D_PLANE;
• inline float get_theta ( float *v, int idx)
• N_allocated__bytes + ( S ) ; ⁇
• N_outstanding_allocated_bytes + ( S ) ; ⁇
• N_allocated_bytes_high_water_mark ⁇
• strcpy (Largest_single_mem_request_fname, FILE ); ⁇
• N_allocated_bytes + (S); ⁇
• strcp (Largest_single_mem_request_fname, FILE ); ⁇
• MPI_Abort MPI_COMM_WORLD , ABORT_INVALID_IDX ) ;
• MPI_Abort MPI_COMM_WORLD , ABORT_INVALID_IDX ) ;
• MPI_Abort MPI_COMM_WORLD , ABORT_FAILED_MALLOC
• strcpy (Largest_single_mem_request_fname, FILE );
• N_allocated_bytes + (S) ⁇
• N_outstanding_allocated_bytes + ( S ) ; ⁇ if (N_outstanding__allocated_bytes ⁇
• N_allocated_bytes_high_water_mark ⁇
• strcpy (Largest_single_mem_jrequest_fname, FILE ); ⁇
• N_allocated_bytes + (S); ⁇
• strcpy (Largest_single_mem_request_fname, FILE ); ⁇
• N_allocated_bytes + ( S ) ; ⁇
• N_outstanding_allocated_bytes + (S); ⁇ if (N_outstanding_allocated_bytes
• strcpy (Largest_single_meiti_request_fname, FILE .);
• MPI_Abort MPI_COMM_WORLD , ABORT_FAILED_MALLOC ) ;
• strcpy (Largest_single_mem_request_fname, FILE _)
• N_freed_b tes + Allocs [ aid ] . len ;

Landscapes

• Health & Medical Sciences (AREA)
• Life Sciences & Earth Sciences (AREA)
• High Energy & Nuclear Physics (AREA)
• Molecular Biology (AREA)
• Physics & Mathematics (AREA)
• General Physics & Mathematics (AREA)
• Spectroscopy & Molecular Physics (AREA)
• Engineering & Computer Science (AREA)
• Medical Informatics (AREA)
• Optics & Photonics (AREA)
• Heart & Thoracic Surgery (AREA)
• Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
• Microelectronics & Electronic Packaging (AREA)
• Pathology (AREA)
• Radiology & Medical Imaging (AREA)
• Biomedical Technology (AREA)
• Biophysics (AREA)
• Surgery (AREA)
• Animal Behavior & Ethology (AREA)
• General Health & Medical Sciences (AREA)
• Public Health (AREA)
• Veterinary Medicine (AREA)
• Apparatus For Radiation Diagnosis (AREA)
• Measurement Of Radiation (AREA)
• Analysing Materials By The Use Of Radiation (AREA)

Priority Applications (4)

Applications Claiming Priority (2)

Related Child Applications (2)

Publications (2)

Family

ID=45372011

Family Applications (1)

Country Status (4)

Cited By (8)

Families Citing this family (8)

Family Cites Families (9)

• 2011
  • 2011-04-04 EP EP11798535.8A patent/EP2553493B1/en active Active
  • 2011-04-04 WO PCT/US2011/031104 patent/WO2011162851A2/en not_active Ceased
  • 2011-04-04 JP JP2013502916A patent/JP2013527918A/ja active Pending
  • 2011-04-04 US US13/638,314 patent/US8766180B2/en active Active
• 2014
  • 2014-09-05 JP JP2014180864A patent/JP2014232119A/ja active Pending

Non-Patent Citations (1)

Also Published As

Similar Documents

Legal Events